Are We Putting Too Many Eggs in One Basket with Cloudflare?

November 18, 2025 – Today the internet learned a brutal lesson: never put all your eggs in one anycast basket.

From social media giants to AI powerhouses, music streaming, e-commerce, gaming, and even the very site that tracks outages (Downdetector), millions of users worldwide were met with HTTP 500 errors, endless loading spinners, and the dreaded Internal server error on Cloudflare’s network message.

Even the irony was on Cloudflare: Downdetector.com, the world’s most popular outage-tracking site, was itself unreachable because it routes through Cloudflare.

The Roll Call of the Fallen – A Partial List of Websites Using Cloudflare That Went Down Today

At approximately 11:48 UTC, Cloudflare, the invisible shield protecting roughly 20–25% of all websites on the planet, suffered a catastrophic internal outage that rippled across the globe within minutes. For nearly two hours, every website using Cloudflare (in proxied/orange-cloud mode) either returned 500/503 errors, showed Error 500: Internal Server Error pages, or simply timed out. 

This is not exaggeration. These are real-time confirmed victims from the November 18 outage:

• X.com (formerly Twitter) – Completely inaccessible in many regions
• ChatGPT / OpenAI.com – Internal Server Error for millions
• Discord.com – Voice and text services degraded worldwide
• Spotify.com – Playback stopped, app and web player down
• Canva.com – Editor completely unreachable, projects not loading
• Perplexity.ai – Queries failing, service unavailable
• Gemini.google.com – AI responses blocked, timeout errors
• Letterboxd.com – Site down, “500 Internal Server Error” widespread
• Leagueoflegends.com – Login servers offline, game client failing
• Grindr.com – App and website inaccessible for most users
• Shopify.com – Storefronts and admin panels throwing Cloudflare errors
• Coinbase.com – Trading platform down, wallet access disrupted
• Authy.com – 2FA service unreachable, blocking logins everywhere
• Paypal.com – Payments and checkout failing globally
• Claude.ai (Anthropic) – Chatbot returning errors, service degraded
• Doordash.com – Ordering system down, app not loading
• Uber.com – Ride booking unavailable in multiple countries
• Depop.com – Marketplace and app completely inaccessible
• Runescape.com – Game worlds offline, login queue broken
• Downdetector.com – The outage tracker itself ironically bricked by Cloudflare
• Feedly, Notion, Medium, Canva, Crunchbase, Gumroad – Productivity apocalypse
• Thousands of Shopify stores – E-commerce bloodbath on a random Tuesday
• Major news sites – The Guardian (parts), Politico, Quartz, many regional papers
• Gaming platforms – League of Legends login, Riot Games, Grinder, Roblox (partial)
• VPN dashboards, crypto wallets, banking portals

If a site showed the infamous Cloudflare Attention Required! Cloudflare challenge page or a stark white 5xx error today, congratulations, it was one of the millions of websites using Cloudflare.

Why This Keeps Happening

Cloudflare is brilliant. It’s free for basic use, insanely performant, and blocks more malicious traffic than most countries’ militaries. But when you turn on that little orange cloud in your DNS settings, you are handing your entire internet presence to a single company in San Francisco.

Today proved, once again, that all the websites using Cloudflare rise together… and fall together.

This is the third global Cloudflare outage in 18 months, severe enough to trend worldwide. Previous incidents (June 2022, July 2024, and now November 2025) follow the same pattern: an internal deployment or routing issue cascades because of Cloudflare’s anycast architecture – the very thing that makes it fast also makes it a single point of failure for everyone.

There Is Another Way

You do not have to gamble your entire online business on a third-party proxy you don’t control.

Enterprises are quietly contacting sovereign edge platforms providers, private, custom-built Cloudflare replacements that deliver the same (or better) performance, security, and global reach but owned and operated by you, not rented from San Francisco.

No more waking up to find your fate decided by someone else’s incident dashboard.

The Bottom Line

Today, all the websites using Cloudflare were reminded of a simple truth:

Renting your uptime is convenient… until it isn’t.

When Cloudflare goes down, the internet doesn’t break. Just the part of it that trusted a single company too much.

The smart ones won’t let there be a next time.

If you’re tired of being collateral damage every time Cloudflare hiccups, it’s time to talk about building the infrastructure you should have owned from day one.

Because the next outage isn’t a question of if.

It’s a question of when and whether your business will still be standing when it happens.